Super-Resolution Microscopy Approaches for Live Cell Imaging

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Biophysical Journal Volume 107 October 2014 1777–1784 1777 Biophysical Review

Super-resolution Microscopy Approaches for Live Cell Imaging

Antoine G. Godin,1,2 Brahim Lounis,1,2 and Laurent Cognet1,2,* 1University of Bordeaux, LP2N, UMR 5298, Talence, France; and 2Institut d’Optique Graduate School and Centre National de la Recherche Scientiﬁque, LP2N, UMR 5298, Talence, France

ABSTRACT By delivering optical images with spatial resolutions below the diffraction limit, several super-resolution ﬂuores- cence microscopy techniques opened new opportunities to study biological structures with details approaching molecular structure sizes. They have now become methods of choice for imaging proteins and their nanoscale dynamic organizations in live cells. In this mini-review, we describe and compare the main far-ﬁeld super-resolution approaches that allow studying endogenous or overexpressed proteins in live cells.

INTRODUCTION The decryption of cell functions and subcellular processes has using saturable optical processes that deexcite emitters constantly benefited from advances in microscopy. In partic- formerly excited by a focused laser beam. These processes ular, the developments of fluorescence microscopy and of work to prevent fluorescence emission from specific regions numerous fluorescent probes allowing the study of specific of the excitation beam by driving the molecules in these biomolecules at work in their native environment were instru- regions between bright and dark states using a depletion light mental to the advance of live cell mechanism investigations. beam. One elegant and efficient strategy consists of using The optical resolution of microscopes is limited by the diffrac- stimulated emission by a high-intensity (>MW/cm2), tion of light, which commonly sets a limit of ~l/2 in far-field doughnut-shaped laser beam superimposed with the focused microscopy. By delivering optical images with spatial resolu- excitation laser beam, completely preventing fluorescence tions below the diffraction limit, super-resolution fluorescence emission from emitters in peripheral regions of the excitation microscopy offered new promises to study molecular pro- beam. This process was coined ‘‘stimulated emission deple- cesses with greater detail than with conventional microscopies tion’’ (STED) (3). A doughnut-shaped depletion beam is the (1,2). Most of these methods rely on the control of the number simplest design; however, in general, any depletion beam of emitting molecules in specific imaging volumes. This can featuring a spatial intensity distribution with one or several be achieved by controlling local emitter fluorescent state intensity zeroes can be used to perform STED images. populations or the labeling densities of fluorescing probes at To generate a super-resolved image with STED based on any given time during the image acquisition process. In this local excitation volumes, one must scan the excitation/deple- mini-review, we will discuss the key features of super-resolution effective volumes over the sample in a deterministic tion techniques used for live-cell studies. We schematically point-by-point manner or by use of parallelized scanning divide them into three major groups: those based on highly schemes (4,5). STED was successfully applied in several localized fluorescence emission volumes; those based on live samples to study slow morphing and movements of structured illumination; and those based on single-molecule organelles such as reticulum endoplasmic or microtubules localizations. A didactic representation of the three families (6), subcellular organization in live cells (7), and synaptic of super-resolution approaches is presented in Fig. 1. structures in live samples (7–9). For live cell studies, one should bear in mind that relatively high laser powers are needed in STED, especially when using continuous wave SUPER-RESOLUTION BASED ON HIGHLY laser beams (e.g., ~MW/cm2 (10)). Using pulsed excitation LOCALIZED FLUORESCENCE EMISSION beam together with time-gating detection allowed a ~2–3- VOLUMES fold reduction in laser power (11). In addition, photobleach- Stimulated emission depletion (STED) and ing is a limiting factor for long-term live sample imaging reversible saturable optical fluorescence because each fluorescent molecule undergoes a large num- transition (RESOLFT) ber of exciting/de-exciting cycles in the depletion beam. An approach similar to STED using much lower inten- In a far-field confocal microscope, the effective fluorescence sities to deplete emitting molecular levels (~kW/cm2)(12) volume can be reduced below the diffraction limit (3)by is based on reversible photoswitching of marker proteins between a fluorescence-activated and a nonactivated state Submitted June 18, 2014, and accepted for publication August 7, 2014. (13–15), whereby one of the transitions is accomplished *Correspondence: [email protected] by means of a spatial intensity distribution featuring a Editor: Brian Salzberg. Ó 2014 by the Biophysical Society 0006-3495/14/10/1777/8 $2.00 http://dx.doi.org/10.1016/j.bpj.2014.08.028 1778 Godin et al.

C RESOLFT/STED A Object

Excitation over the sample Depletion Scanning laser beams D Structured illumination microscopy (SIM) 3 patterns*3 modulations = 9 images Software

B Diffraction Reconstruction limited image E Single molecule approaches Acquired Images

...... Image Reconstruction from localizations Single molecule localization

FIGURE 1 Schematic description of the superresolution microscopy approaches. All images for this didactic description are computer-generated. Object to be imaged consisted of fluorescent emitters (A) and corresponding diffraction-limited image (B). (C) In RESOLFT/STED, a focused excitation beam (cyan) superimposed with a doughnut-shaped depletion beam (red) are scanned over the sample to acquire an image at high resolution (down to ~50– 80 nm in live cells). (D) In SIM, after the required software reconstruction, multiple wide-field images are acquired using sinusoidal illumination grid patterns to obtain high-resolution images (down to ~50–100 nm in live cells using nonlinear saturated illumination). (E) In single-molecule localization microscopy, a large number of wide-field images containing a few isolated single fluorescent emitters are successively acquired. A high-resolution image is reconstructed from the localizations of each individual molecule. Resolutions down to ~50 nm are commonly achieved in live cells. In the example provided, we considered the detection of 80% of the molecules present in the object image. Scale bar represents 1 mm. To see this figure in color, go online. zero. This generalized approach was named after ‘‘revers- vides only an approximately twofold resolution enhance- ible saturable optical fluorescence transition’’ (RESOLFT). ment of standard wide-field microscopy as compared to Bright photostable switchable fluorophores and fluorescent other super-resolution methods (19). Nonlinear saturated proteins development were particularly instrumental in the SIM using fluorophore saturation or photoswitchable pro- development of these techniques (14–16). Importantly, fluo- teins as in RESOLFT can achieve higher resolution rescent proteins provide specific 1:1 protein labeling and enhancement (~50 nm), but requires an increased number offer the possibility of intracellular live cell imaging. of image acquisitions (up to 63) and a complex reconstruction process (20,21). SIM has been demonstrated for long- term, live cell imaging in microtubules and other dynamic STRUCTURED ILLUMINATION MICROSCOPY (SIM) structures (21–23). Three-dimensional SIM imaging has Structured illumination microscopy (SIM) is based on stan- been further achieved using 15 different pattern acquisitions dard wide-field microscopy and is compatible with most per axial planes for reconstruction instead of nine images to standard fluorophores and labeling protocols. It uses reject the out-of-focus light (24). Whole-cell volume imag- nonuniform illuminations with known spatial patterns ing has been performed using three-dimensional SIM in two (e.g., originally a sinusoidal grid, but other illumination dis- colors (25). And, interestingly, fast SIM imaging (11 Hz) tributions can also be used (17)). From multiple acquisitions has even been developed with a 100-nm resolution for a (e.g., nine images, incorporating three phase shifts for three small field of view (~8 8 mm2)(18). pattern orientations (18)), high spatial frequency information is retrieved with a dedicated algorithm, comprising a SINGLE-MOLECULE LOCALIZATION method inaccessible with standard illumination schemes MICROSCOPY APPROACHES (19). Contrary to standard laser scanning modalities like STED/RESOLFT, SIM allows acquisition of a large field It is well known that the position of isolated single fluores- of view over limited times. However, SIM routinely pro- cent emitters can be determined by image analysis with

Biophysical Journal 107(8) 1777–1784 Live Cell Super-resolution Imaging 1779 greater precision than is available from the diffraction limit with the use of reducing/oxidizing buffers that can affect alone. This feature, which has been used for more than 20 cell integrity (41). Of special interest is that STORM has years in live cell, single-particle/-molecule studies (26), is been shown to take advantage of some reduction in thiol key to providing today’s super-resolved images. Super-res- glutathione, which is naturally present at millimolar con- olution methods based on single-molecule localizations centrations in bacteria (42) or in specific cell compartments simply consist of reconstructing an image from single mole- such as the nuclei of eukaryote cells when using buffers with cule localizations retrieved from a large number of movie low cellular toxicity (43,44). Organelles from live cells such frames (typically thousands of camera frames). The main as the membrane and mitochondria have also been investi- requirement is that each frame contains the detection of gated using multicolor STORM (45–47). spatially well-separated fluorescent emitters (27). Inasmuch Analogously to dSTORM, where stochastic photoswitch- as a large volume of single-molecule detections must to be ing is used to control the number of emitting fluorophores, registered to reconstruct a high-content super-resolved im- ground state depletion microscopy followed by individual age, this acquisition process is inherently slow (typically molecular return (GSDIM) covers the techniques employ- more than a few seconds). Below, we describe three families ing the transition between the fluorescent singlet state and of such approaches, distinguishing how fluorescent mole- the metastable triplet state as a stochastic on-off switch cules are stochastically isolated from nonfluorescent ones (48,49). More precisely, efficient transition to the long-lived in each camera frame. triplet state is achieved in such techniques by using high- excitation intensities combined with an imaging buffer, similar to STORM, to allow obtaining triplet lifetimes just Photoactivation localization microscopy (PALM) long enough to leave only a few emitting fluorophores at The development of photoactivation localization micro- any time in each image. Under these conditions, GSDIM scopy (PALM) (28,29) is closely linked to the advent of pho- has been employed for imaging living cells using both toactivatable proteins (16), which allows us to control, by fluorescent protein tags (as in PALM) or various organic light, the density of fluorescing proteins in each image. fluorophores (as in STORM) that selectively bind to tagged Although it is restricted to expression systems, preventing proteins (50,51). the study of endogenous proteins in their native environment, PALM takes advantage of the versatility and speci- Universal point accumulation imaging in the ficity of genetically encoded, fluorescently tagged nanoscale topography (uPAINT) molecules in cells, and has quickly become the tool of choice for super-resolution live cell imaging. PALM, by In contrast with PALM, STORM, and GSDIM, which are design, is not restricted to biomolecules present at the based on the emitters stochastic photoswitching, the method cell plasma membrane, and allows the study of intracellular known as ‘‘universal point accumulation imaging in the biomolecules. By tracking the movement of each indi- nanoscale topography’’ (uPAINT) (52–55) captures real- vidual protein, PALM also allows measuring local diffusion time molecular interactions to control the density of fluores- properties in living cells on short timescales (30–32) and cent emitters suitable for single molecule identification in cellular structural changes in three dimensions on longer each image. In the uPAINT approach, target molecules are timescales (33). individually imaged when a specific ligand coupled to a fluorescent dye binds to the target molecule. Unbound ligands freely diffuse in the imaging buffer Stochastic optical reconstruction microscopy (with typical diffusion constants of approximately tens of (STORM) and ground state depletion microscopy mm2/s) and, due to an oblique illumination excitation, are followed by individual molecular return (GSDIM) not excited in an efficient manner. Therefore, unbound Stochastic optical reconstruction microscopy (STORM) (34) ligands are not detected efficiently by a detector operating and direct STORM (dSTORM) (35) use switchable organic at a typical video rate, in contrast to bound ligands, which fluorophores placed in specific buffers (e.g., with reducing diffuse together with a membrane receptor (with typical properties) instead of using fluorescent proteins as in diffusion constants of <1 mm2/s) in the oblique laser illumi- PALM. Those probes can be targeted on genetically encoded nation. With uPAINT, any binding entity conjugated to flu- or endogenous proteins using adequate ligands. STORM was orophores having high specificity toward a target molecule first demonstrated using Cy3-Cy5 pairs (34) but was quickly (e.g., natural/synthetic ligand, antibody) can be used as extended in dSTORM to other synthetic fluorophores that fluorescent probes to reveal the targeted molecules. Appli- can be stochastically and reversibly switched in the imaging cations include receptors and GPI-anchored proteins diffus- buffers (36,37). STORM is particularly powerful for fixed ing on live cell membranes labeled with antibodies or cells applications (38–40), and can be extended in three synthetic ligands. uPAINT was also applied to image and dimensions (38). One caution is that live cell experiments track endogenous receptors such as glutamate receptors should be performed with great care due to possible issues in neurons (54) and epidermal growth factor receptors

Biophysical Journal 107(8) 1777–1784 1780 Godin et al.

FIGURE 2 Examples of achievements obtained A B STED SIM with superresolution microscopy in live biological samples. (A) STED: continuous-wave STED images of the yellow fluorescent protein (citrine) targeted to the endoplasmic reticulum in live cells revealing small tubules (~60 nm). Image sequences show morphing of the endoplasmic reticulum at arrows (pixel size ¼ 20 nm, 10 s recording time per image). Scale bar ¼ 1 mm. This figure was adapted from Hein et al. (6). (B) SIM: total-internal 1 μm 2 μm reflection microscopy image series of eGFP-a- tubulin in a live S2 cell and corresponding SIM C PALM D STORM E uPAINT images revealing the elongation followed by a rapid shrinking of a microtubule. Integration time EGFR dimers of 270 ms per frame. This figure was adapted from Kner et al. (18). (C) PALM: numerous single 1 μm trajectories of b3-integrin fused with mEOS2, 1 μm obtained on a single MEF cell with PALM, revealing that b3-integrin undergo slower free- diffusion inside focal adhesions (gray) than Diffusive outside, as well as confined diffusion and immobi- Confined lization. Figure adapted from Rossier et al. (31). 1 μm Immobile 5 μm (D) STORM: spatial dynamics of cortical actin skeleton stained with Lifeact-HaloTag/ATTO655. Each reconstruction was obtained using 1000 frames (2 ms per frame). Scale bar ¼ 1 mm. This figure was adapted from Wilmes et al. (47). (E) uPAINT: live cell superresolution imaging of membrane epidermal growth factor receptor (EGFR) dimers based on single-molecule fluorescence resonance energy transfer induced by fluorescent ligand activation. (Inset) Preferential cell-edge localization of EGFR dimers. In addition, uPAINT provides numerous single- molecule trajectories on a single cell, allowing the extraction of the diffusion properties of the EGFR dimer population from the whole-ligand-activated EGFR population. This figure was adapted from Winckler et al. (56). To see this figure in color, go online.

(EGFRs) at high densities in culture living cells (56). Com- molecular (re)organizations. In particular, to grasp the full parisons between transfected nonendogenous receptors and spectrum of mobility behaviors of biomolecules (up to 1 endogenous glutamate receptors were also performed (57). mm2/s for membrane receptors), fast video rate acquisition A similar approach allowed tracking and imaging by is required (20–100 Hz) on large fields of view. Yet, in all continuously labeling sodium ion channels in live cells, us- super-resolution methods, breaking the diffraction limit on ing small fluorescently labeled molecule agents that display a given field of view comes at the expense of the acquisition reversible binding to the sodium ion channel (58). Interest- time. ingly, combining single-molecule fluorescence resonance In point scanning RESOLFT/STED methods, a compro- energy transfer and dual-color uPAINT allowed the specific mise between imaging large fields of view and fast acquisi- super-resolution imaging and tracking of interacting recep- tion speed has to be made because of the requirement for a tors activated by their cognate ligand in live cells (56). This dense pixilation. Imaging rates in RESOLFT are rather slow feature stems from the fact that in this uPAINT study, because pixel integration times are limited by the protein fluorescently tagged ligands are directly used as imaging photoswitching processes. Being based on stimulated emis- probes, allowing us to extract and image the population of sion, STED is not subject to this fundamental limit. How- activated functional receptors upon ligand binding in real- ever, increased resolution being achieved with high laser time. Examples of high resolution images obtained based powers, care should be taken to ensure live cell integrity. on highly localized fluorescence emission volumes, struc- RESOLFT/STED methods are able to resolve the move- tured illumination microscopy and single-molecule locali- ments of the slow structures (typically 10–30 s for 512 zation microscopy are presented in Fig. 2 . 512 pixels) such as, for instance, microtubule networks organizations (22,59) and neuron morphology dynamics (60,61), in cell cultures and live animals. Interestingly, by DISCUSSION acquiring small fields of view, the dynamics of nanoscale structures can be monitored with higher imaging rates Dynamics and resolution in live cells (28 Hz for 60 82 pixels) (62). Recently, large paralleli- In live cells or organisms, supramolecular structures and zation of the depletion patterns combined with the use of organelles morph often in reaction to stimuli (seconds to matrix detectors drastically increased imaging speed over minutes and up to hours) over relatively short timescales. large field of view in RESOLFT (59) and STED (5). To fully understand the cell signaling that induces those pro- In its standard form, SIM uses nine wide-field fluorescent cesses, it is important to investigate therein the dynamics of images to build a super-resolved image in typically 1 s.

Biophysical Journal 107(8) 1777–1784 Live Cell Super-resolution Imaging 1781

Many applications of SIM were described for imaging sub- uPAINT, STED, and to some extent, STORM/GSDIM, cellular structures in living cells (18,23,25). With saturated should be considered, inasmuch as they are compatible SIM (the structured illumination approach giving better with studying endogenous receptors in living cells. Interest- spatial resolution), additional illumination patterns are ingly, by synchronizing single-molecule detection and needed for the reconstruction (21), leading to degraded ligand-induced receptor activation, uPAINT is, to date, the time resolution. Important to note for SIM techniques is only super-resolution method that allows studying, in real- that any aberration, sample movement, or fluorophore pho- time, specifically activated functional receptors and their tobleaching during the image sequence will induce artifacts interactions at the membrane of living cells (56). that will strongly affect the quality of reconstructed super- Of note, protein number quantification can in principle resolved images (18). be performed using PALM and uPAINT, inasmuch as, in In single-molecule localization approaches, two time- these methods, photobleaching irreversibly turns off fluoro- scales are relevant: phores after their detection. In STORM/GSDIM, however, reversible stochastic switching of fluorophores can bias 1. Raw-images acquisition rate. This sets the individual such quantitative analysis because observing the same fluo- molecules’ tracking time resolution (1–10 of ms). This rescent molecules more than once is plausible. Interestingly, rate also sets the single-molecule pointing accuracy recent advances in self-labeling proteins such as the SNAP, through its impact on the signal/noise of each molecular CLIP, and Halo tags, allow efficient live cell protein label- detection (63,64). The analysis of single-molecule trajec- ing, including intracellular ones (46,51,67), and provide tories provides local mobility maps on live cell regions valuable tools to perform multicolor super-resolution imag- with high spatiotemporal resolutions (30–32,54,57) ing that can, in principle, be applied to STED, STORM, or 2. Total number of images needed to reconstruct a super- uPAINT. resolved image (which sets the rate). Indeed, in addition to the pointing accuracy, the local density of single- molecule detections obtained from a studied structure Studying structures in three dimensions also plays a central role in the final spatial resolution Several methods were developed to improve axial resolution (as announced by the Nyquist theorem). For instance, in fluorescence imaging. The most widely used strategies to obtain images with 10-nm resolution, local densities are based on single-molecule localization and provide axial m 2 of at least 10,000 detection/ m are needed. Hence, information (~20–70 nm) of the position without severely thousands of image frames are commonly acquired, justi- altering either the radial or time precision. By shaping the fying global recording times of approximately seconds detection point-spread function along the axial position, to minutes. This timescale directly defines the time reso- single-molecule position can be precisely determined along lution at which nanoscale organization of molecular the optical axis (38,68,69). Detecting molecules using two assemblies (e.g.. cellular organelles) can be analyzed. objectives (70,71), by moving the sample in the axial direc- tion (72) or by interfering the signals obtained from two Computer analysis requirements objectives (73,74), could also yield similar resolution improvements along the optical axis. Although the last Conversely to RESOLFT/STED methods that do not require approach is the most precise (~20 nm), it is also the most any postacquisition analysis, the main source of SIM’s complex to implement. Finally, STED/RESOLFT can also complexity lies in the sophisticated algorithms required be extended in three dimensions by scanning distinct axial for image reconstruction. As in single-molecule localiza- planes independently (9). tion-based techniques, positions of the emitting single molecules must be retrieved using cutting-edge software (described in a recent comparative study (65)). SUMMARY Super-resolution approaches described in this mini-review were proven to deliver information on subcellular organiza- Labeling strategies and consequences tion at different timescales using various labeling probes. Expression systems in concert with fluorescent protein Table 1 summarizes the main pros/cons of the approaches engineering provide a method of choice to study, with discussed here and outlines the different spatiotemporal high specificity, subcellular organizations in live cells, fundamental limits. making RESOLFT, SIM, and PALM essential methods for the applications described in this review. However, despite CONCLUSION their wide applicability, one should bear in mind that some signaling and structural artifacts can arise due to the During the last two decades, super-resolution approaches use of fluorescent proteins (66). In this context, using fluo- have provided new insights into subcellular organization rophores conjugated to specific ligands, the methods SIM, at nanometer-scale resolutions. Several of these methods

Biophysical Journal 107(8) 1777–1784 1782 Godin et al.

TABLE 1 Comparison of the superresolution approaches presented in this review Single-molecule approaches Approach STED RESOLFT SIM PALM STORM/GSDIM uPAINT Resolution (live samples) 50–70 nm 80–100 nm 50c–100 nm 50 nm 50 nm 50 nm Toxicity þþ þþþ þþþ þþþ þ þþþ Endogenous þþþ — þþþ — þþþ þþþ Time for single image (a field of 50 50 mm 10–100 s ~0.1 sa >500 s ~ 3 sb ~1 s > 2s > 2s > 2s is considered for comparison) Intracellular labeling (live) þ þþþ þþþ þþþ þ — Implementation complexity þ þ þ þþþ þþþ þþþ Reconstruction algorithm N.A.a N.A.b þþþþþþþ Dynamics of large molecular structures þþþ þþþ þþþ þþ þþ þþ Dynamics of single molecule þ þ þ þþþ þþ þþþ Multicolor imaging þ þþ þþþ þþ þþþ þþþ aParallelization of STED nanoscopy using optical lattices was recently achieved with an imaging rate of 12.5 Hz for a 2.9 2.9 mm image (5). It requires a simple reconstruction algorithm. bLarge parallelization of the depletion patterns drastically increased imaging speed over a larger field of view in RESOLFT (120 100 mmin~3s(59)). It requires a simple reconstruction algorithm. cUsing nonlinear saturated SIM allows accuracy enhancement up to 50 nm (21). have reached a sufficient level of maturity to make them 5. Yang, B., F. Przybilla, ., B. Lounis. 2014. Large parallelization of routinely applicable to many biological systems. As des- STED nanoscopy using optical lattices. Opt. Express. 22:5581–5589. cribed in this mini-review, each approach has its own advan- 6. Hein, B., K. I. Willig, and S. W. Hell. 2008. Stimulated emission depletion (STED) nanoscopy of a fluorescent protein-labeled organelle tages and drawbacks for live cell imaging. However, in inside a living cell. Proc. Natl. Acad. Sci. USA. 105:14271–14276. combining these techniques, one can tackle specific bio- 7. Westphal, V., S. O. Rizzoli, ., S. W. Hell. 2008. Video-rate far-field logical questions to take advantage of the strengths of optical nanoscopy dissects synaptic vesicle movement. Science. 320: each method, such as the insurance of orthogonal labeling 246–249. . for multicolor imaging (47). 8. Na¨gerl, U. V., K. I. Willig, , T. Bonhoeffer. 2008. Live-cell imaging of dendritic spines by STED microscopy. Proc. Natl. Acad. Sci. USA. Other recent approaches can also be implemented to in- 105:18982–18987. crease the image resolution by a factor of 2–4 upon applying 9. Tønnesen, J., G. Katona, .,U.V.Na¨gerl. 2014. Spine neck plasti- software analytical tools to standard experimental imaging city regulates compartmentalization of synapses. Nat. Neurosci. 17: data sets. These include deconvolution microscopy (75), 678–685. . stochastic optical fluctuation imaging (76), Bayesian local- 10. Willig, K. I., B. Harke, , S. W. Hell. 2007. STED microscopy with continuous wave beams. Nat. Methods. 4:915–918. ization microscopy (77), and compressed sensing (78–80). 11. Vicidomini, G., G. Moneron, ., S. W. Hell. 2011. Sharper low-power We foresee that correlating the information gathered by STED nanoscopy by time gating. Nat. Methods. 8:571–573. such specific molecular imaging methods with structural 12. Testa, I., N. T. Urban, ., S. W. Hell. 2012. Nanoscopy of living brain information obtained by electron microscopy will provide slices with low light levels. Neuron. 75:992–1000. new insights into molecular subcellular organization and 13. Hofmann, M., C. Eggeling, ., S. W. Hell. 2005. Breaking the diffrac- interaction. tion barrier in fluorescence microscopy at low light intensities by using reversibly photoswitchable proteins. Proc. Natl. Acad. Sci. USA. 102:17565–17569. We acknowledge financial support from the Fondation pour la Recherche . Medicale, Fonds de Recherche Nature et Technologies du Quebec (to 14. Dedecker, P., J. Hotta, , J. Hofkens. 2007. Subdiffraction imaging through the selective donut-mode depletion of thermally stable photo- A.G.G.), Agence Nationale de la Recherche, Region Aquitaine, the French switchable fluorophores: numerical analysis and application to the fluo- Ministry of Education and Research, the Centre National de la Research rescent protein Dronpa. J. Am. Chem. Soc. 129:16132–16141. Scientifique, the European Research Council, and FranceBioImaging (grant 15. Grotjohann, T., I. Testa, ., S. W. Hell. 2011. Diffraction-unlimited No. ANR-10-INSB-04-01). all-optical imaging and writing with a photochromic GFP. Nature. 478:204–208. 16. Patterson, G. H., and J. Lippincott-Schwartz. 2002. A photoactivatable REFERENCES GFP for selective photolabeling of proteins and cells. Science. 297: 1873–1877. 1. Hell, S. W. 2007. Far-field optical nanoscopy. Science. 316:1153–1158. 17. Gao, L., L. Shao, ., E. Betzig. 2012. Noninvasive imaging beyond the 2. Huang, B., M. Bates, and X. Zhuang. 2009. Super-resolution fluores- diffraction limit of 3D dynamics in thickly fluorescent specimens. Cell. cence microscopy. Annu. Rev. Biochem. 78:993–1016. 151:1370–1385. 18. Kner, P., B. B. Chhun, ., M. G. Gustafsson. 2009. Super-resolu- 3. Hell, S. W., and J. Wichmann. 1994. Breaking the diffraction resolution tion video microscopy of live cells by structured illumination. Nat. limit by stimulated emission: stimulated-emission-depletion fluores- Methods. 6:339–342. cence microscopy. Opt. Lett. 19:780–782. 19. Gustafsson, M. G. 2000. Surpassing the lateral resolution limit by a 4. Bingen, P., M. Reuss, ., S. W. Hell. 2011. Parallelized STED fluores- factor of two using structured illumination microscopy. J. Microsc. cence nanoscopy. Opt. Express. 19:23716–23726. 198:82–87.

Biophysical Journal 107(8) 1777–1784 Live Cell Super-resolution Imaging 1783

20. Gustafsson, M. G. 2005. Nonlinear structured-illumination micro- 42. Wang, W., G. W. Li, ., X. Zhuang. 2011. Chromosome organization scopy: wide-field fluorescence imaging with theoretically unlimited by a nucleoid-associated protein in live bacteria. Science. 333:1445– resolution. Proc. Natl. Acad. Sci. USA. 102:13081–13086. 1449. 21. Rego, E. H., L. Shao, ., M. G. Gustafsson. 2012. Nonlinear struc- 43. Klein, T., A. Lo¨schberger, ., M. Sauer. 2011. Live-cell dSTORM with tured-illumination microscopy with a photoswitchable protein reveals SNAP-tag fusion proteins. Nat. Methods. 8:7–9. cellular structures at 50-nm resolution. Proc. Natl. Acad. Sci. USA. 44. Benke, A., and S. Manley. 2012. Live-cell dSTORM of cellular DNA 109:E135–E143. based on direct DNA labeling. ChemBioChem. 13:298–301. . 22. York, A. G., S. H. Parekh, , H. Shroff. 2012. Resolution doubling in 45. Shim, S. H., C. Xia, ., X. Zhuang. 2012. Super-resolution fluores- live, multicellular organisms via multifocal structured illumination cence imaging of organelles in live cells with photoswitchable mem- microscopy. Nat. Methods. 9:749–754. brane probes. Proc. Natl. Acad. Sci. USA. 109:13978–13983. 23. Olshausen, P. V., H. J. Defeu Soufo, ., A. Rohrbach. 2013. Super- 46. Benke, A., N. Olivier, ., S. Manley. 2012. Multicolor single molecule resolution imaging of dynamic MreB filaments in B. subtilis—a multi- tracking of stochastically active synthetic dyes. Nano Lett. 12:2619– ple-motor-driven transport? Biophys. J. 105:1171–1181. 2624. 24. Shao, L., P. Kner, ., M. G. Gustafsson. 2011. Super-resolution 3D 47. Wilmes, S., M. Staufenbiel, ., J. Piehler. 2012. Triple-color super- microscopy of live whole cells using structured illumination. Nat. resolution imaging of live cells: resolving submicroscopic receptor or- Methods. 8:1044–1046. ganization in the plasma membrane. Angew. Chem. Int. Ed. Engl. 25. Fiolka, R., L. Shao, ., M. G. Gustafsson. 2012. Time-lapse two-color 51:4868–4871. 3D imaging of live cells with doubled resolution using structured illu- 48. Fo¨lling, J., M. Bossi, ., S. W. Hell. 2008. Fluorescence nanoscopy mination. Proc. Natl. Acad. Sci. USA. 109:5311–5315. by ground-state depletion and single-molecule return. Nat. Methods. 26. Cognet, L., C. Leduc, and B. Lounis. 2014. Advances in live-cell 5:943–945. single-particle tracking and dynamic super-resolution imaging. Curr. 49. Testa, I., C. A. Wurm, ., C. Eggeling. 2010. Multicolor fluorescence Opin. Chem. Biol. 20:78–85. nanoscopy in fixed and living cells by exciting conventional fluorophores with a single wavelength. Biophys. J. 99:2686–2694. 27. Betzig, E. 1995. Proposed method for molecular optical imaging. Opt. Lett. 20:237–239. 50. Lalkens, B., I. Testa, ., S. W. Hell. 2012. MRT letter: Nanoscopy of protein colocalization in living cells by STED and GSDIM. Microsc. 28. Betzig, E., G. H. Patterson, ., H. F. Hess. 2006. Imaging intracellular Res. Tech. 75:1–6. fluorescent proteins at nanometer resolution. Science. 313:1642–1645. 51. Lukinavicius, G., K. Umezawa, ., K. Johnsson. 2013. A near-infrared 29. Hess, S. T., T. P. Girirajan, and M. D. Mason. 2006. Ultra-high resolu- fluorophore for live-cell super-resolution microscopy of cellular pro- tion imaging by fluorescence photoactivation localization microscopy. teins. Nat. Chem. 5:132–139. Biophys. J. 91:4258–4272. 52. Sharonov, A., and R. M. Hochstrasser. 2006. Wide-field subdiffraction . 30. Manley, S., J. M. Gillette, , J. Lippincott-Schwartz. 2008. High- imaging by accumulated binding of diffusing probes. Proc. Natl. Acad. density mapping of single-molecule trajectories with photoactivated Sci. USA. 103:18911–18916. localization microscopy. Nat. Methods. 5:155–157. 53. Cognet, L., D. A. Tsyboulski, and R. B. Weisman. 2008. Subdiffraction . b 31. Rossier, O., V. Octeau, , G. Giannone. 2012. Integrins 1 and far-field imaging of luminescent single-walled carbon nanotubes. Nano b 3 exhibit distinct dynamic nanoscale organizations inside focal adhe- Lett. 8:749–753. sions. Nat. Cell Biol. 14:1057–1067. 54. Giannone, G., E. Hosy, ., L. Cognet. 2010. Dynamic super-resolution 32. Gudheti, M. V., N. M. Curthoys, ., S. T. Hess. 2013. Actin mediates imaging of endogenous proteins on living cells at ultra-high density. the nanoscale membrane organization of the clustered membrane Biophys. J. 99:1303–1310. protein influenza hemagglutinin. Biophys. J. 104:2182–2192. 55. Giannone, G., E. Hosy, ., L. Cognet. 2013. High-content super-reso- 33. Fu, G., T. Huang, ., J. Xiao. 2010. In vivo structure of the E. coli lution imaging of live cell by uPAINT. Methods Mol. Biol. 950:95–110. FtsZ-ring revealed by photoactivated localization microscopy (PALM). 56. Winckler, P., L. Lartigue, ., L. Cognet. 2013. Identification and super- PLoS ONE. 5:e12682. resolution imaging of ligand-activated receptor dimers in live cells. Sci 34. Rust, M. J., M. Bates, and X. Zhuang. 2006. Sub-diffraction-limit Rep. 3:2387. imaging by stochastic optical reconstruction microscopy (STORM). 57. Nair, D., E. Hosy, ., J. B. Sibarita. 2013. Super-resolution imaging Nat. Methods. 3:793–795. reveals that AMPA receptors inside synapses are dynamically orga- 35. Wombacher, R., M. Heidbreder, ., M. Sauer. 2010. Live-cell super- nized in nanodomains regulated by PSD95. J. Neurosci. 33:13204– resolution imaging with trimethoprim conjugates. Nat. Methods. 7: 13224. 717–719. 58. Ondrus, A. E., H. L. Lee, ., J. Du Bois. 2012. Fluorescent saxitoxins 36. Heilemann, M., S. van de Linde, ., M. Sauer. 2009. Super-resolution for live cell imaging of single voltage-gated sodium ion channels imaging with small organic fluorophores. Angew. Chem. Int. Ed. Engl. beyond the optical diffraction limit. Chem. Biol. 19:902–912. 48:6903–6908. 59. Chmyrov, A., J. Keller, ., S. W. Hell. 2013. Nanoscopy with more 37. Schermelleh, L., R. Heintzmann, and H. Leonhardt. 2010. A guide to than 100,000 ‘doughnuts’. Nat. Methods. 10:737–740. super-resolution fluorescence microscopy. J. Cell Biol. 190:165–175. 60. Rochefort, N. L., and A. Konnerth. 2012. Dendritic spines: from struc- 38. Huang, B., W. Wang, ., X. Zhuang. 2008. Three-dimensional super- ture to in vivo function. EMBO Rep. 13:699–708. resolution imaging by stochastic optical reconstruction microscopy. 61. Berning, S., K. I. Willig, ., S. W. Hell. 2012. Nanoscopy in a living Science. 319:810–813. mouse brain. Science. 335:551. 39. Dani, A., B. Huang, ., X. Zhuang. 2010. Super-resolution imaging of 62. Lauterbach, M. A., J. Keller, ., S. W. Hell. 2010. Comparing video- chemical synapses in the brain. Neuron. 68:843–856. rate STED nanoscopy and confocal microscopy of living neurons. 40. Lo¨schberger, A., S. van de Linde, ., M. Sauer. 2012. Super-resolution J Biophotonics. 3:417–424. imaging visualizes the eightfold symmetry of gp210 proteins around 63. Bobroff, N. 1986. Position measurement with a resolution and noise- the nuclear pore complex and resolves the central channel with nano- limited instrument. Rev. Sci. Instrum. 57:1152–1157. meter resolution. J. Cell Sci. 125:570–575. 64. Thompson, R. E., D. R. Larson, and W. W. Webb. 2002. Precise 41. Jones, S. A., S. H. Shim, ., X. Zhuang. 2011. Fast, three-dimensional nanometer localization analysis for individual fluorescent probes. super-resolution imaging of live cells. Nat. Methods. 8:499–508. Biophys. J. 82:2775–2783.

Biophysical Journal 107(8) 1777–1784 1784 Godin et al.

65. Chenouard, N., I. Smal, ., E. Meijering. 2014. Objective comparison 73. Shtengel, G., J. A. Galbraith, ., H. F. Hess. 2009. Interferometric fluo- of particle tracking methods. Nat. Methods. 11:281–289. rescent super-resolution microscopy resolves 3D cellular ultrastruc- 66. Swulius, M. T., and G. J. Jensen. 2012. The helical MreB cytoskeleton ture. Proc. Natl. Acad. Sci. USA. 106:3125–3130. in Escherichia coli MC1000/pLE7 is an artifact of the N-Terminal 74. Kanchanawong, P., G. Shtengel, ., C. M. Waterman. 2010. yellow fluorescent protein tag. J. Bacteriol. 194:6382–6386. Nanoscale architecture of integrin-based cell adhesions. Nature. 468: 67. Stagge, F., G. Y. Mitronova, ., S. Jakobs. 2013. SNAP-, CLIP- and 580–584. Halo-tag labelling of budding yeast cells. PLoS ONE. 8:e78745. 75. Sibarita, J. B. 2005. Deconvolution microscopy. Adv. Biochem. Eng. 68. Pavani, S. R., M. A. Thompson, ., W. E. Moerner. 2009. Three- Biotechnol. 95:201–243. dimensional, single-molecule fluorescence imaging beyond the diffrac- 76. Dertinger, T., R. Colyer, ., J. Enderlein. 2009. Fast, background-free, tion limit by using a double-helix point spread function. Proc. Natl. 3D super-resolution optical fluctuation imaging (SOFI). Proc. Natl. Acad. Sci. USA. 106:2995–2999. Acad. Sci. USA. 106:22287–22292. . 69. Izeddin, I., M. El Beheiry, , M. Dahan. 2012. PSF shaping using . adaptive optics for three-dimensional single-molecule super-resolution 77. Cox, S., E. Rosten, , R. Heintzmann. 2012. Bayesian localization imaging and tracking. Opt. Express. 20:4957–4967. microscopy reveals nanoscale podosome dynamics. Nat. Methods. 9:195–200. 70. Juette, M. F., T. J. Gould, ., J. Bewersdorf. 2008. Three-dimensional . sub-100 nm resolution fluorescence microscopy of thick samples. Nat. 78. Studer, V., J. Bobin, , M. Dahan. 2012. Compressive fluorescence Methods. 5:527–529. microscopy for biological and hyperspectral imaging. Proc. Natl. Acad. Sci. USA. 109:E1679–E1687. 71. Xu, K., H. P. Babcock, and X. Zhuang. 2012. Dual-objective STORM reveals three-dimensional filament organization in the actin cytoskel- 79. Zhu, L., W. Zhang, ., B. Huang. 2012. Faster STORM using com- eton. Nat. Methods. 9:185–188. pressed sensing. Nat. Methods. 9:721–723. 72. Chasles, F., B. Dubertret, and A. C. Boccara. 2006. Full-field optical 80. Zanella, R., G. Zanghirati, ., G. Vicidomini. 2013. Towards real-time sectioning and three-dimensional localization of fluorescent particles image deconvolution: application to confocal and STED microscopy. using focal plane modulation. Opt. Lett. 31:1274–1276. Sci. Rep. 3:2523.

Biophysical Journal 107(8) 1777–1784